INDEX

Explanations

the word "other" followed by some context

oai_token-act-pair · gpt-3.5-turbo

references to "other" entities or subjects in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ffee

-0.67

 Fuji

-0.64

ister

-0.60

-0.59

ensibly

-0.57

 Bride

-0.57

itan

-0.57

hod

-0.56

POSITIVE LOGITS

worldly

1.58

etheless

0.95

wise

0.93

swer

0.85

quickShipAvailable

0.83

 kinds

0.77

 aspects

0.73

wcs

0.71

parts

0.70

types

0.70

Activations Density 0.078%

the word "other" followed by some context

references to "other" entities or subjects in various contexts

No Comments

No Known Activations